1) Introduction

This appeal raised a novel and increasingly practical question for the criminal law: whether “gold pieces” (the in‑game currency of Old School RuneScape) constitute “property” within s.4(1) Theft Act 1968, such that they may be stolen.

The respondent, a Jagex employee, was alleged to have gained unauthorised access to 68 player accounts, stripped them of approximately 705 billion gold pieces, and then sold the gold pieces off‑line for fiat currency and Bitcoin. Count 2 alleged theft of gold pieces (valued at £543,123) “belonging to Jagex Ltd”. Counts 1 and 3–5 were linked (Computer Misuse Act and POCA money laundering) and depended, in substance, on the existence of a predicate “theft” offence.

At a preparatory hearing (under the approach in R v Quillan [2015] 1 WLR 4673), the judge dismissed the Theft Act foundation by ruling that gold pieces were not “property”, treating them as non‑rivalrous and akin to “pure information”. The prosecution appealed.

2) Summary of the Judgment

The Court of Appeal allowed the prosecution’s appeal and answered the preliminary issue in the prosecution’s favour: gold pieces within Old School RuneScape are “property” capable of being stolen.

The court held that the judge’s rivalrousness analysis was wrong; and, more importantly, that the Theft Act’s concept of “property” is not controlled by private-law property doctrines or by the contractual position between Jagex and players. The correct focus is the statutory language and the criminal-law purpose: preventing dishonest taking of things that can sensibly be described as capable of being stolen, subject to limited exceptions (notably “pure information” per Oxford v Moss (1979) 68 Cr App R 182).

3) Analysis

3.1 Precedents Cited (and How They Shaped the Decision)

A. Framing “property” by context

• Nokes v Doncaster Amalgamated Colliers Ltd [1940] AC 1014: the court adopted the contextual approach—“property” is not a fixed term of art; it takes meaning from statutory purpose. This supported treating Theft Act “property” as a criminal-law concept aimed at dishonest taking, not a mirror of civil ownership.
• Sir Roy Goode KC, “What is Property” (2023) LQR 139 1: cited to acknowledge the conceptual complexity of “property”, reinforcing the court’s decision to avoid abstract general theory and focus on the Theft Act’s function.

B. Civil-law indicia: helpful but not determinative

• National Provincial Bank v Ainsworth [1965] AC 1175: the four Ainsworth criteria (definable, identifiable, assumable, stable) were treated as satisfied by gold pieces. However, the court stressed that Ainsworth addresses private-law property recognition; it does not control Theft Act interpretation. The decision thus uses Ainsworth as supporting confirmation rather than a gateway test.
• Tulip Trading Ltd v Bitcoin Association BSV & others [2023] EWCA Civ 83 [2023] 4 WLR 16 and AA v Persons unknown [2019] EWHC 3556 [2020] 4 WLR 35: relied on principally to illustrate modern judicial understanding that digital assets may be “property” and to emphasise the need to distinguish the functional asset from the underlying code. The court drew on Birss LJ’s “all there is, is software” observation to explain why physical tangibility is not required.

C. The criminal-law boundary: “pure information” and why this case is different

• Oxford v Moss (1979) 68 Cr App R 182: treated as the central “exception” case—“pure information” (knowledge as such) is not Theft Act property. The court refined the rationale: knowledge exists in the mind and cannot be “stolen” in the requisite sense because the “victim” is not deprived of the knowledge by another’s acquisition.
• United Aircraft Corporation v Federal Commissioner of Taxation (1943) 68 CLR 525: used to support the proposition that “knowledge” is not property in a legal sense; secrecy does not transmute knowledge into property. This underpinned the court’s explanation of why “pure information” is excluded.

D. “Widest ambit” of Theft Act property; and indicia for intangible property

• Attorney General of Hong Kong v Nai-Keung [1987] 1 WLR 1339: pivotal. The Privy Council held that export quotas were “other intangible property” and stressed the definition’s “widest ambit”. The Court of Appeal extracted two practical indicia: (i) the thing is bought and sold; and (ii) it is susceptible to dishonest dealing depriving the holder of its benefits. RuneScape gold pieces fit both.
• R v Smith (Michaael Andrew) [2011] EWCA Crim 66: used to show that criminal-law “property” can extend to things (like illegal drugs) even where civil-law enforcement of rights would be barred by public policy. This supported the court’s key proposition: Theft Act property does not depend on enforceable civil title.

E. Criminal law can diverge from civil law; avoid importing civil complexities

• R v Hinks [2001] 2 AC 241 (and the discussion of Lawrence [1972] AC 626 and Gomez [1993] AC 442): deployed to rebut the defence instinct that Theft Act concepts must track private-law validity. The court relied on Lord Steyn’s recognition that “disharmony” can exist, and that criminal law may properly reach conduct not actionable civilly, to avoid placing dishonest takings “beyond reach”.
• Ivey v Genting Casinos (UK) Ltd (t/a Crockfords Club) [2017] UKSC 67 [2018] AC 391: cited as an example where divergence would be undesirable (dishonesty), to contrast with “property” where divergence is tolerable.
• R v Woodman [1974] QB 754, R v Turner (No 2) [1971] 1 WLR 901, R v Kelly [1999] QB 621: cited to support that s.5 Theft Act “belonging to another” includes factual possession/control and is not limited by civil legality. This reinforced the court’s view that s.4 should not be trapped by civil ownership analysis.

F. Comparative decisions (persuasive context, not binding)

• HR, 31 januari 2012, NJ 2012, 536 m.nt. Keijzer (Neth.): a Runescape virtual-item theft decision considered supportive in “common sense” terms; the Dutch court treated in-game items as having independent existence and value, and treated civil ownership by the developer as irrelevant to criminal appropriation.
• R v Dixon [2015] NZSC 147 (and Kenyon v Spry [2008] HCA 56): cited and ultimately treated as of limited assistance because of its statute-specific, tangibility-leaning reasoning about files; but it reinforced the contextual meaning of “property”.
• Yeates (a pseudonym) v The King [2025] VSCA 288: a post-hearing Australian decision holding Bitcoin is property; helpful for analysis distinguishing “mere information” from rivalrous digital assets and differentiating asset from code. The Court of Appeal found it useful, while declining to make private-law tests determinative for Theft Act purposes.

G. Statutory landscape and law reform materials

• The court engaged with the Law Commission’s 2023 and 2024 reports and the concept of “rivalrousness”, but emphasised their focus on private-law property rights rather than Theft Act property.
• The court noted the Property (Digital Assets) Act 2025 as a private-law clarification that “a thing … is not prevented from being the object of personal property rights” merely because it is neither a thing in possession nor a thing in action—again, not directly determinative of the Theft Act question.
• Other criminal-statute references (illustrative of contextual drafting): R v Whitely (1991) 93 Cr App R 25 (criminal damage and computer data); POCA definitional provisions discussed but treated as irrelevant to Part 7 money laundering counts in this appeal.

3.2 Legal Reasoning

A. The correct statutory focus: s.4 Theft Act in its criminal purpose

The court’s organising move was to relocate the inquiry from civil “ownership” to Theft Act function. Section 4 provides an inclusive, broad definition (“money… things in action and other intangible property”). In the court’s view, that breadth—reinforced by s.5 (possession/control)—means the criminal law should treat as “property” anything that (in ordinary language and social reality) can be stolen, unless a principled exception applies.

B. The main exception: “pure information” is not property

The court accepted the orthodox boundary that “pure information” (knowledge as such) is not Theft Act property, explaining this as a matter of (i) ontology (it exists in the mind) and (ii) deprivation (acquisition by another does not deprive the holder).

Gold pieces were not “knowledge”: they were functional, account-specific, exclusive-use digital assets that can be transferred and whose loss deprives the account holder of gameplay utility and of market value in practice.

C. Distinguishing asset from data/code

A central analytical clarification was that the relevant “thing” is the functional in-game asset (gold pieces), not the underlying instantiated data as data. Data can be copied; but the relevant asset is the exclusive, system-recognised instance that confers control and value within that ecosystem. The court analogised with cryptocurrency analysis: the public code is not owned, but the token/value unit is treated as property.

D. Rivalrousness: (i) not necessary as a legal test; (ii) satisfied on the facts anyway

The court rejected the preparatory-hearing judge’s conclusion that gold pieces are non-rivalrous because they are fungible and “infinite” in supply. It held that “infinite possible supply” does not negate rivalrousness of particular instances held and controlled at a given time—just as paper clips can be indefinitely manufactured but each clip is still rivalrous property.

Moreover, even if rivalrousness is treated as a useful indicium (as in modern digital-asset private law discussions), gold pieces are rivalrous because one player’s use/transfer of specified gold pieces necessarily excludes others from using those same gold pieces.

E. Contract terms (EULA/T&Cs) are not determinative

Jagex’s contractual documents sought to characterise virtual currency as not the player’s “private property” and to reserve extensive powers to delete/restrict currency. The court held that such arrangements may affect private-law enforceability, but do not settle the criminal-law meaning of “property”. Criminal law can treat things as stealable even when civil law would deny enforceable proprietary remedies.

F. “Belonging to another” can attach via control/possession without civil ownership

The court also rejected the attempt to narrow the issue to “Jagex’s property” only. Under s.5, property may “belong” to more than one person: Jagex may have possession/control (servers and ultimate powers), while players may have sufficient factual control in use and transfer. This flexibility prevents overly technical pleading from defeating Theft Act protection.

3.3 Impact

A. Immediate doctrinal effect

• Establishes (at appellate level) that in-game virtual currency with real functionality and market value may be “other intangible property” for s.4 Theft Act 1968.
• Clarifies that “property” for Theft Act purposes is not confined by private-law ownership, contractual disclaimers, or civil enforceability—subject to limited carve-outs like “pure information”.
• Rejects a simplistic “infinite supply implies non-property” approach, and corrects the rivalrousness analysis where the asset is exclusive-use within a system.

B. Practical prosecutorial consequences

• Enables theft charges (and therefore related money laundering counts premised on theft proceeds) for hacking-based appropriation of in-game currency and potentially other in-game assets, where the conduct deprives players of use/value.
• Reduces the risk of a “charging lacuna” where Computer Misuse Act offences capture access but not the economic gravamen of appropriation and conversion.

C. Likely future litigation questions

• How far the reasoning extends to other in-game assets (skins, weapons, accounts) with different transferability rules, scarcity mechanics, or purely cosmetic function.
• The boundary between “pure information” and “functional digital assets”, especially where the “asset” is essentially access rights, credentials, or purely informational advantages.
• The relationship between Theft Act “property” and regulatory regimes: the court rejected the argument that recognising property for theft would automatically trigger FCA/AML obligations for game developers, but defendants may continue to raise such policy arguments.

4) Complex Concepts Simplified

• “Other intangible property” (s.4 Theft Act): non-physical things that can still count as property—provided they are not excluded by principle (like “pure information”) or by specific statutory/common-law carve-outs.
• “Pure information”: knowledge as such (e.g., knowing exam questions). It is not property for theft because it exists in the mind and taking it does not deprive the “owner” of their knowledge.
• Rivalrousness: a thing is rivalrous if one person’s use/consumption necessarily excludes another from using/consuming that same thing. “Unlimited possible supply” does not mean a specific instance is not rivalrous.
• Asset vs code/data: copying the data representing an asset does not necessarily copy the asset’s functional existence within its system. The court treated the “gold piece” as the system-recognised, exclusive-use unit, not the abstract data string.
• “Belonging to another” (s.5 Theft Act): property can “belong” to someone who has possession or control, even without civil-law ownership. Multiple people can qualify simultaneously (e.g., platform operator and player).
• Why civil ownership is not decisive: criminal law protects public order by penalising dishonest taking; it need not coincide perfectly with civil remedies or contractual risk-allocation. The court used R v Hinks [2001] 2 AC 241 to illustrate permissible civil/criminal “disharmony”.

5) Conclusion

Lakeman, R. v establishes a clear criminal-law rule: in-game gold pieces in Old School RuneScape are “property” within s.4 Theft Act 1968 and can be stolen. The court reached that result by emphasising the Theft Act’s broad, purposive concept of property, confining the “pure information” exclusion to knowledge, and refusing to let private-law ownership analysis or contractual labels dictate the scope of criminal theft.

The decision is significant beyond gaming: it is a modern appellate statement that the criminal law can recognise functional digital assets as stealable property where they are exclusive-use, value-bearing units susceptible to dishonest deprivation—without waiting for perfect alignment with civil proprietary doctrine.